Light-weight routines for processing sky survey masks
A survey coverage mask is made up of a large number of tiles forming a mosaic on the sky. Each tile can be represented as a group of convex spherical polygons. We will optimize for the case when the polygons are small, less than 1 degree, and the mosaic covers a large fraction of the sky. So we can reference the polygons efficiently by a center point and sort them with spatial indices.
The main operations to be carried out are:
- query polygons that contain a given point
- draw random samples of points from the area covered
Check out the python notebooks in the demos/ subdirectory. They may be viewed online: http://nbviewer.jupyter.org/github/bengranett/minimask/tree/master/demos/
Load a mask polygon file and generate uniform random samples:
>>> M = minimask.Mask("survey.txt.gz")
>>> ra, dec = M.sample(n=10000) # generate 10,000 points
>>> ra, dec = M.sample(density=10) # ...or with number density 10 / square degreeCheck if points are inside a polygon:
>>> inside = M.contains(ra, dec)
>>> print inside
[ True True True ..., True True True]Query weights:
>>> w = M.get_combined_weight(ra, dec, operation='sum')Make plots:
>>> healpy.mollview()
>>> for x,y in M.render():
>>> healpy.projplot(x, y, lonlat=True)
>>> pylab.show()
Pixelize with healpix:
>>> map = M.pixelize(nside=64)
>>> healpy.mollview(map)
Ben Granett, Dida Markovic